RE: 2 more days25 May 2021 12:31
SR, it is in a research paper by students of investment studies emailed to me so no link. Most of the paper is statistical modelling based on historical price movements and really make no sense in bits. The gist is summed up by the conclusion bits. I have cut and paste it here so I hope it helps the BB. I didn't make it up:
Conclusion
In this research the results of the MSCI Minimum Volatility index rebalances between 2011 and 2016 are
presented. The main finding is that stocks that increase in weight return an average of +1.58% between
announcement and effective day. Stocks that are (partly) sold during the index rebalance return an
average of -0.63% in same time period. This results in an index spread (difference between buy and sell
basket) of +2.21%, making the rebalance 2.21% more expensive on effective day than on announcement
day. Investors lose every rebalance (10% turnover) 0.22% and with two rebalances per year this
accumulates to a performance drag of 0.44% annually. Additionally the following conclusions can be
drawn:
1. The index effect is persistent, of structural nature and statistically significant. Only one rebalance
(November 2011) of the 12 rebalances investigated yields a marginal negative spread between
AD and ED. On effective day, all but one (2015-II) rebalances show positive spread returns.
Furthermore, no single rebalance result dominates the overall average.
2. The spread between AD and ED doubles to +4.5% when constructing baskets on basis of liquidity
(less liquid stocks get a higher weight). This indicates that the index effect is caused by market
impact.
3. The index effect found is statistically significant with significance at 5% level found for event
windows on announcement day (AD), between announcement day and effective day (AD+1 –
ED-1) and on effective day (ED). The reversal effect is mainly observed on the first day after
effective day. For both baskets statistical significant negative spread returns are found. No
significance is found for the flow weighted baskets for the overall time period suggesting that
there is no permanent price change, which adheres to the price pressure hypothesis.